Adopting an external focus (movement effect) during skilled motor actions generates better outcomes than an internal focus (movement itself). This finding is argued to manifest from an external focus more readily activating a common sensory-motor representation that accommodates automatic movement control, as opposed to an internal focus that actively "freezes" large degrees-of-freedom – also referred to as the constrained-action hypothesis. With this in mind, the present study sought to adapt the attention focus framework to contemporary views of limb-target control, where "early" limb-regulation, as opposed to "late" target control, assumes rapid, automatic adjustments. We had participants execute rapid target-directed movements using a digitizing device to translate a cursor on a screen. Meanwhile, participants assumed different focus of attention: external-distal (target), external-proximal (cursor), internal (hand), control (none). There were significantly shorter times to initiate, reach peak velocity and end the movement for the control and external-distal conditions compared to the external-proximal and internal conditions (ps < .05). However, there was significantly smaller terminal error and greater online target adjustments (as indicated by negative relations between the displacements to and after peak velocity) for the external-proximal and internal conditions compared to the external-distal and control conditions (ps < .05). At the same time, there was a significant decline in normalized spatial variability for the external-proximal condition (p < .05). These findings corroborate the constrained-action view, and elaborate on the current consensus by showing that internal/proximal focus more greatly promotes overt adjustments to endpoint responses at the expense of "early" automatic control.